Statistical analyses of the relation between the amount of working hours and impairments to health, based on data from a European survey on working conditions in 2000, clearly reveal that there is a substantial correlation between the number of working hours per week and the frequencies of health complaints. This applies to both musculo-skeletal disorders as well as to psycho-vegetative complaints. The relationship of the duration of the exposure to working conditions to health impairments is moderated by a great number of individual (e.g., age) and situational (e.g., shift-work) variables, showing additive or interactive effects for which selected examples have been presented. In general, however, there is a consistent functional relationship between the number or working hours and their effects on the workers that holds over a great variety of conditions. It is argued that requests for extending working hours should thus be handled with care.

Periodic components inherent in actual schedules of flexible working hours and their interference with social rhythms were measured using spectrum analysis. The resulting indicators of periodicity and interference were then related to the reported social impairments of workers. The results show that a suppression of the 24 and the 168 h (seven-day) components (absence of periodicity) in the work schedules predicts reported social impairment. However, even if there are relatively strong 24 and 168 h components left in the work schedules, their interference with the social rhythm (using the phase difference between working hours and the utility of time) further predicts impairment. The results thus indicate that the periodicity of working hours and the amount of (social) desynchronization induced by flexible work schedules can be used both for predicting the impairing effects of the specific work schedules on social well-being as well as for the design of socially acceptable flexible work hours.

Recent studies on flexible working hours show at least some of these working time arrangements seem to be associated with impairing effects of health and well-being. According to available evidence, variability of working hours seems to play an important role. The question, however, is how this variability can be assessed and used to explain or predict impairments. Based on earlier methods used to assess shift-work effects, a time series analysis approach was applied to the matter of flexible working hours. Data on the working hours of 4 week's length of 137 respondents derived from a survey on flexible work hours involving 15 companies of different production and service sectors in Germany were converted to time series and analyzed by spectral analysis. A cluster analysis of the resulting power spectra yielded 5 clusters of flexible work hours. Analyzing these clusters for differences in reported impairments showed that workers who showed suppression of circadian and weekly rhythms experienced severest impairments, especially in circadian controlled functions like sleep and digestion. The results thus indicate that analyzing the periodicity of flexible working hours seems to be a promising approach for predicting impairments which should be investigated further in the future.

No abstract

In Germany, risk assessment of the working time arrangement is legally required, and within this context the authors assessed the usefulness of the fatigue and risk index (FRI) to predict any detrimental effects to health and safety. This assessment was made using data from two different surveys. Each contained records of working hours over a 4-wk period plus information on occupational accidents and health complaints. The independent variables drawn from the FRI parameters included the maxima, means, variances, and index factor scores. Phi-correlations between the FRI (dichotomized to the index of the reference system) and the incidence of an occupational accident were rather moderate, with a maximum correlation of .22 using the mean fatigue index (FI). Correlations between the two index factor scores and health complaints (sleep problems, stomach pain, eructation/heartburn) revealed the FI component, but not the risk index (RI) component, was related to those health complaints. Forward stepwise logistic regression analysis indicated the FI (but not the RI) predicted occupational accidents (FI(factor) odds ratio [OR] = 1.90, confidence interval [CI] 1.23-2.93). When using multiple regression analyses, the FI was able to predict sleep problems and other circadian related-problems, but the regression coefficients were moderate. In general, the results were not considered sufficient to justify a mandatory use of the FRI.